On Wed, Dec 20, 2000 at 10:21:49PM +0100, Michael Schmitz wrote: > > > There's been a big reorganisation. woody is no longer called unstable > > > (sid is) > > > but testing. It's been reverted to the packages from potato and will be > > > filled > > > up with packages from unstable which don't have (too many) critical bugs. > > > I > > > guess it's again a problem of the PPC autobuilder lagging (if there is > > > even > > > any for testing yet?). > > > > OK, this one I -refuse- to take the blame for! :) > > No one tried to blame you ...
Just joking... People have gotten very attached to blaming the autobuilder, and I was amused to see one that wasn't actually its fault. > If testing was just another distribution nickname we should probably set > up another autobuilder for it. As you explained, it isn't. What we > probably need to reconsider is the restriction on testing WRT packages > having no or little release critical bugs in all architectures. Otherwise > testing is going to be useless - I don't know about ARM but the situation > on m68k is currently so bad that I'd not count on things like a new libc > being built and tested fast enough so that it doesn't hold up the other > archs for multiple weeks. We have one autobuilder run by Roman Hodek, and > one fast developer machine run by Christian Steigies, that's all. That's my concern with the design as well. As I see it, it's not meant so much for regular use as for helping release management; for that role, it ought to work excellently. Dan /--------------------------------\ /--------------------------------\ | Daniel Jacobowitz |__| SCS Class of 2002 | | Debian GNU/Linux Developer __ Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \--------------------------------/ \--------------------------------/
